Articles
Clinical Decision Support Systems (3)
Examples of Clinical Decision Support Systems in Hospital:
Automedon/SmartCare–Ventilator Management
This CDSS automates and eVects clinical guidelines for venti-lator management in highdependency environments. Knowledge engineering techniques have been applied to elicit and model the clinical guidelines with expert system method-ologies. These are rule based with forward chaining coupled with temporal reasoning. In addition, software engineering techniques are employed for automatic source code generation.
GIDEON–Diagnosis and Treatment of Infectious Diseases
GIDEON is a Web-based decision support tool, using Bayesian modeling, for the diagnosis of global infectious diseases. GIDEON contains four seamlessly integrated, operational modules. Diagnosis uses clinical signs and symptoms to sug-gest possible infections; epidemiology focuses on the epidemiologic aspects of the disease database; therapy suggests treatment regimens for the disease database; and microbiology can be used for entering laboratory results to assist in the diagnosis process.
LISA–Treatment of Childhood Leukemia
LISA is a Web-based CDSS for providing dosage decision support for the drug regimen in the treatment of acute child-hood lymphoblastic leukemia. The system implements the dose adjustment rules as speciWed in Medical Research Council guidelines. LISA is a rule-based system and uses the PROforma decision engine for determining the course of treatment. Doses have to be constantly monitored, and the treatment regimen is unique for each patient, so this application is ideal for a CDSS.
PERFEX–Expert System for Automated Interpretation of Cardiac Single Photon Emission Computed Tomography Data
PERFEX is a rule-based expert system for automatic interpretation of cardiac SPECT data. This system assists in the diagnosis of coronary artery disease by suggesting the extent and severity of the condition. It provides a patient speciWc summary of the state of the main cardiac arteries together with related information. Knowledge-based methods are used to process and map the 3D visual information into symbolic representations, which are used to assess the extent and severity of cardiovascular disease both quantitatively and qualitatively. The knowledge-based system presents the resulting diagnostic recommendations in both visual and textual forms.
MEDLINE
The simplest form of information retrieval is initiated by a user. MEDLINE is the classic example of an information retrieval resource. That first incarnation of MEDLINE was a bibliographic retrievalsystem for the biomedical literature intended for librarians, called MEDLARS, developed by the National Library of Medicine (NLM).
Medical Subject Headings (MeSH): one might browse the MeSH tree to find terms for retrieving articles from PubMed, or browse therapeutic categories in a drug index, to find antihypertensive medications from a drug formulary, perhaps further indexed.
Decision Support at Brigham and Women’s Hospital
The initial version of CPOE at BWH, implemented in 1993, included relatively little in the way of clinical decision support, as compared to today. The main areas of focus related to reference knowledge, anticipated needs, alerts, reminders, order sets, guidelines, and feedback. An array of reference materials was initially accessible from CD-ROMs. In addition, a bank of calculation tools was compiled to aid clinicians with frequently performed calculations, such as the Cockcroft- Gault equation, for estimated creatinine clearance, and the arterial-alveolar oxygen (A-a) gradient. this below system in hospital work with DSS:
- Medication-related decision support
- Laboratory interventions (Display of charges, Redundant tests, Therapeutic drug monitoring, Tests pending at discharge)
- Radiology interventions (Appropriateness of ordered studies, Critical result tracking)
- Signout: This led to the development of Signout, an application that allows house officers to out their patients electronically. Information including the medication list, key recent laboratory tests, and the code status are abstracted by the information system, and the provider is asked to enter additional data including a problem list and a description of the hospital course
- Assessment of satisfaction with CPOE
- Impact of CPOE on provider time
Maryam Ghalandari